Purpose: Cone-beam CT (CBCT)-based synthetic CTs (sCT) produced with a deep convolutional neural network (DCNN) show high image quality, suggesting their potential usability in adaptive proton therapy workflows. However, the nature of such workflows involving DCNNs prevents the user from having direct control over their output. Therefore, quality control (QC) tools that monitor the sCTs and detect failures or outliers in the generated images are needed. This work evaluates the potential of using a range-probing (RP)-based QC tool to verify sCTs generated by a DCNN. Such a RP QC tool experimentally assesses the CT number accuracy in sCTs.
Methods: A RP QC dataset consisting of repeat CTs (rCT), CBCTs, and RP acquisitions of seven head and neck cancer patients was retrospectively assessed. CBCT-based sCTs were generated using a DCNN. The CT number accuracy in the sCTs was evaluated by computing relative range errors between measured RP fields and RP field simulations based on rCT and sCT images.
Results: Mean relative range errors showed agreement between measured and simulated RP fields, ranging from -1.2% to 1.5% in rCTs, and from -0.7% to 2.7% in sCTs.
Conclusions: The agreement between measured and simulated RP fields suggests the suitability of sCTs for proton dose calculations. This outcome brings sCTs generated by DCNNs closer toward clinical implementation within adaptive proton therapy treatment workflows. The proposed RP QC tool allows for CT number accuracy assessment in sCTs and can provide means of in vivo range verification.
Keywords: adaptive proton therapy; neural networks; proton radiography; quality control; synthetic CT.
© 2021 The Authors. Medical Physics published by Wiley Periodicals LLC on behalf of American Association of Physicists in Medicine.